ALBUQUERQUE, Luiz Carlos de; SILVA, Laercio de Sousa    PARACAMPO, Carla Cristina Paiva. Análise de variáveis que podem interferir no comportamento de seguir regras discrepantes. []. , 22, 1, pp.51-71. ISSN 0188-8145.

^lpt^aCom o objetivo de investigar o controle por regras, 10 universitários foram expostos a um procedimento de escolha de acordo com o modelo; a tarefa era apontar três estímulos de comparação em uma sequência correta. Na Sessão 1 nenhuma sequência era reforçada ou instruída. As contingências da Sessão 2 eram alteradas na Sessão 3, e as contingências da Sessão 3 eram mantidas inalteradas na Sessão 4, iniciada com a regra discrepante. Os participantes eram solicitados a descrever as contingências durante as sessões e foram distribuídos em dois grupos, que diferiam porque a sequência correta na Sessão 2 era estabelecida por reforço diferencial no Grupo RD e por regra no Grupo IN. Os participantes que responderam corretamente na Sessão 3 tenderam a deixar de seguir a regra discrepante na Sessão 4; e os que responderam incorretamente na Sessão 3 tenderam a seguir a regra discrepante na Sessão 4. Um participante, na Sessão 2, formulou uma autorregra antes de o comportamento por ela especificado ter sido estabelecido por suas consequências imediatas. Neste caso, o comportamento especificado pela autorregra permaneceu inalterado nas primeiras tentativas da Sessão 3. Discutem-se os efeitos de perguntas, regras, autorregras, consequências e histórias do ouvinte sobre o comportamento.^len^aAiming to investigate the control of behavior by rules, 10 college students were exposed to a matching to sample procedure with the task of pointing to each of three comparison stimulus in sequence. Each comparison stimulus shared one dimension with the sample: Color (C), Thickness (E) or Form (F).The participants were distributed in two groups. For both groups, in Session 1 no sequence was reinforced or instructed (baseline). During sessions 2, 3, and 4, the correct sequence was reinforced with points exchangeable for money. In Session 2 the sequence programmed as correct was sequence CEF. Initially the correct sequence was reinforced in a continuous reinforcement schedule (CRF), then in fixed ratio 2. The transition from Session 2 to 3 was characterized by the non signalized change in the programmed reinforcement contingencies for the sequences emitted. During Session 3, therefore, there were no instructions and only the sequence EFC was reinforced in CRF. The transition from Session 3 to 4 was characterized by the introduction - in the beginning of Session 4- of a discrepant rule specifying that emitting the sequence FCE the participant would earn points. During Session 4 the sequence EFC was the only one reinforced, in CRF. During Sessions 2, 3, 4, after every third trial the experimenter handled the participant a paper with the following question: "Answer in writing the following question: What is the sequence you have to point to the comparison objects in order to get points?". The answers to the questions were never differentially reinforced by the experimenter. The two groups differed only concerning to the establishment of the correct sequence in Session 2. For the Group RD the correct sequence was established by differential reinforcement and for Group IN by rule. The participants who responded correctly in Session 3 presenting performance sensible to the contingencies were prone to stop following the discrepant rule in Session 4; those who responded incorrectly in Session 3, presenting performance not sensible to the change in contingencies, were prone to follow the discrepant rule in Session 4. One participant from Group RD formulated an auto-rule in Session 2 before the specified behavior had been established by its immediate contingencies. In this case, the behavior specified in the autorule was kept unaltered, e.g., insensible to the contingency change in Session 3 first trials. Considering the results of this study and other similar studies it is suggested that the effects of questions indicating that rules and contingencies are discrepant depend on the joint effect of other variables. The history of the behavior alternative to that specified by the rule should be taken in consideration as affecting the behavior selected by rules and by contingencies. Finally, it is suggested that auto rules may control behavior.

: .

        · | |     · |     · ( pdf )